YAKUTAT BAY GLACIERS 39 wave of advance was extended to other glaciers between 1905 and 1910 was strikingly illustrated by Hidden Glacier, which had advanced two miles and become greatly broken, by Lucia Glacier, which was actively advancing and greatly transformed by crevassing during the summer of 1909, and less strikingly by Hubbard Glacier, whose eastern margin was just beginning to advance in 1909, and by Nunatak Glacier, which had a slight advance in 1910. The third test was that of cessation of advance in those glaciers that had moved forward in 1906; With a sudden, great addition of snow, quickly terminated, and followed by a spasmodic advance of the glaciers thus supplied, it would be expected that the wave of advance would soon die out and the condition of stagnation return. This was also predicted,1 and our observations of 1909-10 show clearly that the prediction was correct, for all of the advancing glaciers of 1906 had returned to stagnation in 1909, and the breaking of the ice had been so healed by ablation that we were once more able to travel over the glaciers, though with far less ease than in 1905. It is believed that the observations of 1909 add the further facts necessary to demonstrate the hypothesis put forward in 1906, and that we are now warranted in stating the explanation with confidence, as an established hypothesis,—a new cause for glacier advance. The sudden forward rush of a glacier, accompanied by pronounced thickening and extensive surface breakage, may be called a glacier flood and the resemblance to a river flood is noteworthy. When heavy rainfall, or unusual melting of snow occurs in the upper reaches of a river, a wave of rising, rapidly down-moving water is started which may cause a flood all along the course, and if a portion of the river is ice covered, the rigid ice crust will be shattered and heaved into a maze of broken ice blocks; but under ordinary conditions the river behaves more normally, slowly rising and falling with variation in supply. So in a glacier, under ordinary conditions, variations in supply manifest themselves in moderate advance or recession; but when a deluge of snow and ice is thrown down in its upper reaches the condition for a spectacular advance, a glacier flood, are introduced. The ice stream flows on more rapidly, its rigid upper and marginal portions are cracked and broken, its surface rises, and its front is pushed forward. There is, however, a striking difference in time occupied by the two classes of floods. A river flood passes from the source to the mouth of the river in a few hours or a few days, and its effects are soon past; but the far less mobile ice requires several years for the transmission of the glacier flood, and its duration is months long, while years are required to bring the ice surface back to its pre-flood state. There is a striking difference, also, in the mode of motion, for the mobile river water itself travels down stream to form the flood while in the glacier flood the advance and breaking of the mass is due to the passage of a wave, not to actual bodily transfer of ice from glacier reservoir to terminus. With these facts in mind it is quite proper to predict the advance of other and longer glaciers in the Yakutat Bay region as a result of avalanching during the 1899 earthquakes. The oscillations of other Alaskan glaciers, for which there is no known climatic warrant, may be due to other earthquakes. Earthquake avalanching may even be responsible for oscillations of mountain glaciers in other regions, as for example in the Himalayas and other youthful, snow-capped mountains still frequently faulted and shaken by seismic disturbances. i Tarr, R. S., The Yakutat Bay Region, Alaska, Professional Paper 64,17. 3. Geol. Survey, 1909. p. 94.